Professional Documents
Culture Documents
Thomas W. Sheehy, et al
Prepared for:
Army Air Mobility Research and Development
Laboratory
January 1976
Reproduced From
Best Available Copy
DISTRIBUTED BY:
&Oor12, Dpr?-T 0
USAAMCDL-TR- 75-43
06 409 1
•January 1976
nal Report
Prepared for
EUSTIS DIRECTORATE
U. S. ARMY AIR MOBILITY RESEARCH AND DEVELOPMENT LAF4ORATORY
Fort Eustis, Va. 23604
NATIONAL TECHNICAL
INFORMATION SERVICE
U O.po.t,,ct
D Cf.-o ,@
VA 2215I
j.....up.
EUSTIS DIflWTOflATE P0!7tTtCN S"AMrTNT
Thes rcprwt has been revic'Mcd by the Eustis Directorate. U. S. Army Air Mobility
Re"torh and Dzwlo~pmeM -Laboratoryand is cor.dcred to be technically sound. The
Purpoae of the program was to dev~elop a proceduret rdc
rdc hh rgo
' rgo zi
t~elicopter rotor aNd hub pylon. Ist
DISCLAIMERS
The* findings in thiesreport are "a't to be construed n Of ficial Deartment of ithe Atmy position unless 6o
a4
desinated by Other ewthorited docuoments.
VW'en Government drawvings, specifications, or Other data are used for any purpose Other than in conne~ction
tovth a definitely related Grovernmewnt procurement operation, the United Sltaes Governmenort thereby incurs no
responsibility nor aity obligation wohatsoeverf and the fact that the Governmente may have formulated, furnished.
or in any way supplied the said dravvings. specifications, or other data is not to be regardel by impolication or
otherwiiseo to in any manner licensing itheoholder or any other person or corwp~raiion. or conveying any rights or
permission, to maisnufacture, use, nr setl any patented invention that maey in ano, way be related thereto.
Trade names cotc-d in this reliort do not constitute an official encdo-jrwrint or approval or. the use of such
connrrororciel hardvwae or software.
DISPOSITION INSTRUCTIONS
Destrov thit report vwhe no longer needed. Do not return it to the originirtor.
.1g
Uncles ifled
USAAflML-TR-75-48[1
4 TITLE ftiod *.)~t~
, a, TV"E of LEP*, a PEICOD COVa"20
Unclassified
Ir. oCkI1aFCATION DOWNGRADING
IS. XRET
WORDS (C..WAN..1t M-,... 84,n1 $0....
0"000- OW 0AW IV60 ..a
A procedure has been developed to determine the contribution of the rotor hub
to the total helicopter drag. The method developed uses a three-dimensional
potential flow analysis to determine the flow environment inwhich the hub
operates combined with empirical data in order to predict the drag of the hub
and its associated interference drag. Predictions using the method are in
good agreement with test data for unfaired and faired rotor hubs.
A review of available rotor hub drag test data was conducted in order to
identify the factors affecting helicopter rotor hub drag. The data base
established was used in the development of the hub drag prediction method
and also to define a systematic wind tunnel ti.st program to refine and
verify the drag prediction method and to investigate in detail the
parameters affecting the dr;ag contribution of the rotor hub.
The results of the data review indicate that the primary factors affecting
the rotor hNo drag and its associated interference drag are the swept
frontal area, the increase In the local dynamic pressure due to the
fuselage/pylon configuration, and the pylon shape.
Unclassified
SECuRITY CLASSIFICATION OF THIS PAGC(WI" Dole EtmdMr
PREFACE
3
TABLE OF CONTENTS
Pae
PREFACE . . . . . . . . . . . . . 3
INTRODUCTION ....................... . . 9
BIBLIOGRAPHY ... . . . . . . . . . . . . . . 43
APPENDIXES
5
Pwoc in page blank
LIST OF TLLUSTRATIONS
Figure
/ /
LIST OF ILLUSTRATIONS (Continued)
7
N
8W
Figure Page
A-6 Required Input Order for Corner Points
Describing an Input Panel ......... 88
B-i Panel Representation of Body Used in
Three-Dimensional Aerodynamic Analysis
Technique (179L) .... .......... 100
B-2 General Arrangement of Lifting Section
Vortex Lattice ietwor't .......... 101
8
7 -, . :7m .. • . . ~.
I
9
HUB DRAG DATA REVIEW
Before discussing the parameters associated with rhe drag of
the rotor hub, it is appropriate that its signifirance be well
established. A sunmary of rotor hub drag trends wi#.* gross
weight i.s shown in Figure 1. rhe geometric compIeex.zjr of the
rotor hub (i.e., articulated, hingeless, faired) is Gaowin to
be a strong influencing factor in determining the contribution
of the hub drag to the total helicopter drag. Figur-e :
demonstrates the magnitude of the hub drag contri.!-iop. The
hub and its associated interFerence drag (i.e., tht: .1rag
created by the alteration of the flow conditions auout a body
due to the proximity of another body) Js responsible for
between 20% and 33% of the total drag of current production
helicopters. Most current helicopter fuselages have not been
designed for high-speed operation, and consequently low drag
was not a primary design ccnsideration. For this reason most
helic7pter fuselages are aerodynamically "unclean"; consequent-
ly, the potential exists for a considerable reduction in the
fuselage drag itself. 1 The line in Figure 2 also demonstrates
the penalty paid for the rotor bib drag if the drag of the
fuselage aloi.e is reduced by 50% without a proportionate
reduction in rotor hub drag. It is apparent from Figure 2
that every effort should be expended in reducing the drag
contribution of the rotor hub and that it is imperative that
the drag of the hub be accurately predicted.
The hub drag data presented herein reprEo it the data
supporting the major conclusions drawn. 4The results of the
hub drag data review are identified in greatr detail in
References 1 and 2. The data identified by circle symbols in
the figures represent cata drawn from internal Sikorsky Air-
craft and United Technologies Research Center reports. For
the purpose of clarity and since most of the reports are not
r -nerally available, the individual iata points are not
iaentified by a reference; however, the bibliography includes
the titles and authors of the reports from which the data
were obtained.
10
WAY
up
area related to the rotor hub refers to the "swept" frontal
area which represents the total projected frontal area of a
rotating rotor hub. Figure 3 shows the trend of rotor hub
drag with hub swept frontal area for a range of unfaired rotor
hubs. Vie rotor hub drag (fil) represents only the drag of the
hub itseAf and does not include interference effects. The
data sh-own in Figure 3 indicate that approximately 1 square
foot of dral can be saved for each squate foot reduction in
frontal ate&. It should he pointed out that the scatter in
this figurcý and in other figures concerned with hub frontal
area ran be partially attributed to errors in estimating the
value of hub frontal area. Thia parameter was siLply not
identified in some reports, and frontal areas weze measured
from sketches or photographs.
The effect of including interference drag associated with the
rotor hub is demonstrated in Figure 4, which shows the varia-
tion in rotor hub incremental drag with frontal area. The
rotor hub incremental drag ('rj) represents the difference
in drag of the helicopter configuration with the hub and the
configuration without the hub, and consequently includes tCe
interference drag.
Although no consistent variation with hub frontal area is
apparent, there is a general increase in incremental hub drag
with increaaing frontal area. The curve in Figure 3 has been
duplicated in Figure 4 to demonstrate the potentially large
differences between hub drag (fil) and rotor hub incremezi.tal
drag (TH). The values obtained from the hub drag curve have
also been increased by 25% in Figure 4 to simulate a 25% in-
crease in the dynamic pressure such as might be encountered
at the hub location on a fuselage/pylon configuration. It is
apparent that not even a 25% increase in the local dynamic
pressure can account for the difference in rotor hub f 1 ! and
fH.
The scatter in the incremental hub drag values and the large
differences between flI and ?I are indicative of the magnitude
of the hub interference drag and imply that the interference
drag is dependent on the individual fuselage/pylon configura-
tion tested.
11
ellipsoid fairing is deam±ned to fully enclose the rotor hub
in an effo:t to make the hub configuration a more streamlined
aerodynamic shape.
12
rods) area may offset the interference drag reduction. Be- 4
cause of this trade-off between interference drag and drag due
to increajsed frontal area, there is an optimum hub/pylon
separation distance which will yield the minimum incremental
hub drag for a given hub/pylon/fuselage configuration.
Fuselage Angle of Attack
The data reviewed indicate that the incremental hub drag of
faired or unfaired hubs is not significantly affected by
changes in the fuselage angle of attack over a relatively wide
range of body attituder. This is demonstrated in Figure 8,
which represents only a part of the data supporting this con-
clusion. The reason the hub incremental drag is not affected
by body attitude appears to be that the presence of the fuse-
lage/pylon configuration tends to suppress changes in the flow
conditions, such as local dynamic piessure and flow direction
at the hub location, due to a change in body attitude. It is
interesting to note that this is generally true of test con-
figurations not incorporating a full fuselage simulation al-
though the angle-of-attack range for which this holds is some-
what reduced.
Hub Rotation
The effect of rotating the hub on the hub incremental drag is
shown in Figure 9. Several tests have shown that rotational
effects on unfaired rotor hub incremental drag are negligible.
Wind tunnel tests conducted by Sikorsky also indicate that
blade azimuth position for nonrotating hubs does not signifi-
cantly affect the hub incremental drag. Both of these con-
clusions are based on tests of rotor hubs having four blades
or more. However, for two- or three-bladed hub configurations,
;.t is recommended that tests be conducted with the hub
rotating or with the blades at the 00, 900, 1800 and 2700
azimuth positions.
The data from Reference 3 shown in Figure 9 indicates that the
incremental hub drag increases when the hub is rotated. The
test results from Reference 3 shown were not consistent over
the range of Mc-:h numbers tested, and other tests (such as that
presented in Figure 9) showed little -ariation in hub incre-
mental drag as the hub rotation incraased. Because of the
discrepancies between different test data, it should not be
assumed that rotation does not affect the hub incremental
drag of faired hubs.
. Linville, J. C., AN EXPERIMEITTAL INVESTIGATION OF HIGH-
SPEED ROTORCRAFT DRAG,; Sikorsky Aircraft; USAAMRDL Technical
Repurt 71-46, Eustis Directorate, U. S. Army Air Mobility
R & D Laboratory, Ft. ustis, Virginia, February, 1972,
,•AD740771.
4
13
i!
4
Mach Number
Because the helicopter has traditionally been designed for low
speeds, very little data is available on the effects of Mach
number on rotor hub drag. An exception is the data presented
in Reference 3, which tested a proposed high-speed transport
helicopter configuration over a range of Mach numbers frn. .2
to .6. This test represents a thorough investigaticn of the
effects of Mach number on an unfaired hub, a hub with a rigid
fairing, and an ellipsoidal fairing with and without boundary
layer control (BLC) by tangential blowing. The test results
from this reference, presented in Figure 10, indicate that the
drag of the unfaired hub and the ellipsoidal fairing using
BLC are not significantly affected by Mach numbers between .2
and .4. The incremental drag of the ellipsoidal fairing and
the rigid fairing increases with increasing Mach number over
the range tested. This can probably be attributed to extreme-
ly high local Mach numbers occurring around the bluff fuiring
shape and probably eventually to local shocks. In any case,
the large increases in incremental drag for the hub configura-
tions representing practical hub configurations (i.e., the
unfaired, rigid faired, or ellipsoidal faired hub with BLC)
occur at M > .4, which is not applicable to pure helicopter
configurations, but may be relevant for compound heliccoters.
Scale Effects/Reynolds Number
A large amount of test data has been obtained on the effects
of Reynolds number and on scale effects for streamlined shapes.
Consequently, the scale corrections for test data on these
shapes is relatively well understood. Unfortunately, this is
not the case for bluff bodies and for nonaerodynamic shapes
representative of many helicopter cottponents and in particular
for helicopter rotor hubs.
Results frcm Sikorsky Aircraft tunnel tests indicate that
small-scale tests generally underpredict the drag increments
due to unfaired rotor hubs. However, for faired rotor hubs,
representative hub drag increments can generally be obtained
from small-scale tests. This is illustrated in Figure 11,
which presents test results from full-scale and 1/6-scale
tests of a configuration consisting of a simulated fuselage
upper surface, a pylon, and a faired and an unfaired rotor hub.
14
/
Test data for helicopter configurat.ions indicalte that for the
basic fuselage shape, reasorably small-scale (Reynolds number
>1 million/foot) tests yield accurate drag measu.:.mexats as
discussed in References 4 and 5. The data -resented in these
references did not include comparisons of the configurations
with rotor hubs, however.
Basically, two factors influence the validity of small-scale
test data. The proper flow conditions must be modeled as
closely an possible and the model tested should accurately
represent the geometric characteristics of the actual con-
figuration. The fuselage and pylon region of a helicopter
usually presents no difficulties in this respect provided +he
scale is large enough to obtain Reynolds numbers above sub-
critical for the configuration.
The drag of small-scale rotor hubs appears to fc llow the
basic hub drag trends in Figure 3, which includes some small-
scale data, provided the a'ub geometric properties are
accurately modeled. Recent tests by Sikorsky showed that a
crude geometric model of a rotor hub resulted in only 75% of
the drag measured on an accurate geometric model. It can be
concluded from the data that a true geometric model is a
necessity if an accurate drag estimate is to be obtained.
As far as Reynolds number is concerned, usually a value
sufficient to yield adequate fuselage drag modeling is
sufficient to insure an accurate hub drag. For a rotating hub
the turbulence level is probably adequate to alleviate sub-
critical flow on the small components such as pushrods even
in very small scale tests.
Aerodynamic Sealing
The necessity of good aerodynamic seals in the hub region
depends on the particular hub configuration. Test data
indicate that the ellipsoidal type of fairing must be well
sealed at the juncture of the fairing and the pylon to achieve
any potential drag reduction. On the other hand, the avail-
able data on rigid type fairings indicate that sealing in the
pylon/fairing juncture region is not particularly required.
For unfaired hubs, the pylon cutout (i.e., the opening in the
SSweet, G. E., Julian, L., and Jenkins, J. L., WIND TU14NEL
INVESTIGATION OF THE DRAG AND STATIC STABILITY CHARACTER-
ISTICS OF FOUR HELICOPTER FUSELAGE MODELS, NASA TN D-1363,
National Aeronautics and Space Administration, July 1962.
15
/ .
pylon required for the rotor shaft and controls) can cause
additional drag depending primarily on the shape of the "lip"
on the trailing edge of the cutout. Hoerner 6 presents data
on low drag shapes of holes in surfaces applicable to the
pylon cutout.
As might be expected, the data on both ellipsoidal type
fairings and rigid fairings demonstrated that the openings in
the fairings for the blades should be scaled. This may
present some difficulty for certain types of hubs because the
cutout seals must be moveable or flexible to allow blade
movement, but every effort should be made to seal this region.
Pylon Shape
Generally, the shape of the main rotor pylon has a strong
influence on the incremental hub drag. The pylon influence
manifests itself in two ways. First, the presence of the
pylon (and the fuselage) alters both the local velocity
magnitude and tie direction in the hub region. The local flow
direction will, of course, be parallel to the local surface;
however, the magnitude of the velocity can be considerably
higher than the freestream velocity with consequent increases
in the local dynamic pressure. For many pylon configurations,
the increase in local dynamic pr-.ssure can be the predominant
factor in determining the interference between the pylon and
the hub. For instance, hub incremental drag calculations
based on frontal area and accounting for the increased local
q were reported in Reference 7 to account for the difference
between the hub drag alone and the measured hub incremental
drag for unfaired hubs. Accounting for the local q failed to
account for the drag differences for the faired hubs tested,
however.
The second contribution is through the interference drag. The
flow about the pylon c&n be significantly changed by the
presence of the hub. Assuming that the flow is not initially
separated on the aft pylon, the addition of the rotor hub will
almost certainly induce separation of the flow over the aft
pylon to some extent, thereby increasing its base drag. For
some fuselage/pylon configurations, the effects of the hub
may extend even further -.o influence the aft fuselage region.
Unfortunately, no systematic test data was found which
V Hoerner, Dr. - Ing S. F., FLUID-DYNAMIC DRWG, Second
Edition, Bricktown, New Jersey, Hoerner Fluid Dynamics,
1965.
7 Jenkins, J. L., Winston, M. M., A WIND TUNNEL INVESTIGATION
OF THE LONGITUDINAL AERODYNAMIC CHARACTERISTICS OF TwO FULL
SCALE HELICOPTER FUSELAGE MODELS WITH APPENDAGES, NASA
TN-1364, National Aeronautics and Space Administration,
July 1962.
16
/ lI,
quantified the extent to which the hub interference affected
17
17
HUB DRAG PREDICTION 11ETIIOD
METHOD DEVELOPMENT
Unfaired Hubs
The primary factor influencing the incremental drag of unfaired
hubs is the swept frontal area. The trending equation for the
curve shown in Figure 3 nondimensionalized by hub swept frontal
area is
Ap -. 00057A
2 1
CDH .582 + .0349
CDH p
The CDH obtained from Equation (1) represents the drag of the
hub in free flow and does not include the various interference
effects associated with a hub/pylon combination.
The effect of an increase in the local dynamic pressure in the
hub region is relatively straightforward. If one considers a
pylon/fuselage configuration as shown in Figure 13, the flow
velocity to which the hub is subjected differs from the free-
stream velocity. Since the drag coefficient given by Equation
(1) is based on free-stream dynamic pressure acting on the
hub, CDH must be corrected to account for the local dynamic
pressure. The extent of accelerated or decelerated flow in the
hub region is dependent on the particular configuration in-
volved. However, it is possible that sections of the hub may
be outside the region of influence associated with the pylon/
fuselage induced velocities. To account for that portion of
the hub which may be considered outside of this region of
influence, the ratio of pylon width to hub diameter is used
as a generalizing parameter. The effective drag coefficient
of the rotor hub is then
18
CDI - C• (1 - K2 ) + CDII (1- CPZ) K2 (2)
CDs = AS (1 (3)
AS + A+ t. CDs (1 - K3) + CDs K3 (-Cpz) 3
19
\/
•J/
S~I
As the flow moves around the pylon and fueolage, it accelerates,
causing lcal regions of reduced presture. Following these
regions of low pressure (higher v'.1ocity), the flow deceler-
ates, resulting in a rise in pressure. With such an unfavor-
able pressure gradient the tendency for the boundary layer to
separate from the surface is increased. This susceptability
to flow separation is heightened by the presence of any
disturbance generated by an obstruction such as the rotor hub
or shaft. Consequently, the interference drag can be
evaluated if the strength of the flow disturbance is known and
the stability of the flow on the pylon is gauged from the
pressure gradients on the pylon. A function has been identi-
fied which relates the interference drag to the strength of
the flow disturbance generated by the hub or shaft and the
general flow stability on the pylon. This function is ex-
pressed as
ACDH - K1 (ACpz) ( t/A z) CDa (5)
where Kl - an empirical constant equal to .2
a CpZ - the potential flow pressure differential
(Cpt - Cpz) between the pressure coefficient
at the end of the pylon and the pressure co-
efficient at the location at which the hub is
placed (see Figure 13).
The method of determining CpZ and Cpt will be discussed in a
later section.
CDa in Equation (5) represents the drag coefficient of the
body creating the disturbance. If the shaft height is not
sufficient to remove the hub from the region of increased
"velocity caused by the pylon/fuselage configuration, then
CDa " CDH However, if the shaft height is sufficient, then
the shaft can be consi dered the component creating the dis-
turbance and CDa - CDS.
The relationship given by Equation (5) indicates that as AZ
decreases (i.e., the hub is moved toward the end of the pylon),
the interference drag increases. This is as would be expected
since the boundary layer on the pylon would become more sus-
ceptible to flow separation for a given ACpz due to the
increased boundary layer thickness. The data in Reference 6
indicates that the relationship given by Equation (5) will net
be adequate for hub locations near the end of the pylon (i.e.,
small AZ). This is considered to be irrelevant for helicopter
rotor hub/pylon configurations since this would imply a rear-
ward tilting shaft if the hub were parallel with the local
surface.
20
P*
lip oiI, II III I1101 s I .E
21
particularly blade shanks induce almost total separation on
the fairing afterbody. Based on this, the base drag of the
fairing accounting for the increased velocity due to the
pylon/fuselage is given by
CDFB - (l - CpZ) CpF - (CpC + Cpz) (7)
22
I
I
where Cps - the local pressure coefficient at the 900 and
2700 aziruth position on the hub
CDBS = an empirical drag coefficient determined for the
particular shank configuration
If a shaft or shaft fairing is used in conjunction with the
ellipsoidal fairing, then an additional drag increment must
be included to account for this component. This incremant
is given by
ClS = CDs ( 1 - (CpZ + Cpc)/2) (ASiAr) (10)
Rigid Fairing
23
li . . . -ý t.. 7 V
I I
DF = CDF + CDDS + ACDF (15)
24
AERODYNAMIC ANALYSIS TECIIQIOUE
25
equations satisfying the known flow conditions at the trailing
edge of lifting sections and the requirement that there must
be no flow normal to the surface of the body unless fluid
ejection or suction hc.s boon specified by the user. Once the
singularity strengths have been determined, the flow direction
and magnitude can be computed at the surface control points
and at any arbitrary points located off the body. The proce-
dures for dctermiaiing the singularity strengths are described
in detail in References 10 - 12, and a more detailed description
of the technique can be found in Reference 9. The use of the
three-dimensional aerodynamic analysis computer program
(designated Y179L) is described in Appendix B.
The aerodynamic technique is quite general in application.
The bodies accatable for solution are not required to be
slender, can be lifting or nonlifting, and the body itself
or the flow conditions can be nonsymmetric (i.e., yawed flow).
In addition, the aerodynamic flow field about multiple bodies
can be computed to determine the mutual body interference.
The numerical algorithm used accepts the surface panel de-
scription3 au individual entitities, and, consequently, the
panels may be input in random order and parv ls may be concen-
trated in regions of primary interest which increases the
accuracy of the solution in those regions. This feature is
demonstrated in Figure 16 where the pylon region has been
modeled more accurately than the fuselage surface.
26
The required pressures for predicting the drag of an ellip-I
soidal hub fairing are determined in the same manner as the
values of Cp 7 and Cpl. The ellipsoidal fairing of interest
is geometrically modeled and an analysis performed. The
values of CpC and Cps are obtained directly from the output.
The determination of the equivalent forebody pressuze (CPF) is
not quite as straightforward. Since the fairing is
modeled by a finito number of surface panels, the forebody
of the fairing has an incremental area for each panel on wh4ch
the local pressure is assumed equal to the calculated pressure
for that panel. The resultant equivalent pressure coefficient
(Cpj) acuing on the forebody of the fairing in the drag direc-
tion can then be obtained by numerically integrating these
incremental forces and CpF is given by
11
CPF W 1- 7 Cpi ADi (16)
AF
i- 1
where N - the nunber of panels describing the fairing
forebody (i.e., to the point of maximum
thickness)
Cpi - the local Cp on each panel
ADi - the component of each panel area perpendicular
to the drag direction
The aerodynamic program automatically pei forms the summation
of nondimensionalized forces in the X, Y and Z directions as
well as calculating the moments about each axis. The summa-
tion of forces about the entire body in potential flow without
vorticity, assuming the body description is exact, will be
zero although the distribution of incremental forces will
yield moments whose accuracy depend on the extent of the
viscous flow effects on the particvlar configuration. Conse-
quently, although the potential flow program will inaccurately
predict total pressure recovery on the fairing afterbody, the
CpF determined by the above procedure will be accurate since
the forebody flow is well represented by potential flow.
Geometric Modeling
Although the potential flow solution method is basically the
same for any arbitrary shape, properly modeling the configura-
tion poses a new problem for each body of interest. The
generation of a proper model would normally be a time-consuming
and tedious task for four principal reasons: the numLer of
panels required to model most shapes is large, regions of high
curvature require wore panels than those of low curvature,
panel size should vary smoothly from one adjacent panel to
another, and the total number of panels must be minimized to
27
fit within computing facility constraints of time and size.
To minimize the manual effort in this area, a computerized
method has been developed at Sikorsky Aircraft to complement
the three-dimensional aerodynamic program. The method
developed simplifies the generation of a suitable model for
an arbitrary shape by reducing both the time required to model
the shape and the amount of effort required of the user in
preparing the input.
The body shape is generated by describing relatively few in-
put cross sections by combinations of various types of simple
curves which are then divided into straigLt line increments
based on specific constraints related to the accuracy of the
straight line approximation to the true curve and the basic
requirements of the aerodynamic analysis technique. The
curvature along the input axis (i.e., the axis on which cross
sections are defined) determines the number of cross sections
which are required. In the nose region where the axial
curvature is normally large, several sections may be required,
whereas in the cabin and tail regions, few input sections are
required and the length of individual panels in these rec.tons
is controlled by requesting intermediate cross sections to be
generated by linear interpolation. Because each surface panel
is described as a separate entity, sections of the body may be
generated independently and the resulting output combined to
form the comrlete body.
A description of the Automated Paneling Technique (APT; Deck
Y179A) can be found in Appendix A. The geometry program (APT)
has been succe3sfully used to model configurations ranging
from simple cylindrical shapes to complex compound helicopters
such as the U. S. Army/11ASA/Sikorsky Rotor Systems Research
Aircraft (RSRA). As compared to traditional method.3, the
geometry modeling technique used demonstrates a significant
reduction in time and effort required to model aircraft
configurations.
The surface presstires calculated by the three-dimensional
potential flow program (Y179L) using shapes generated by APT
have been successfully correlated with test data over a wide
range of fuselage configurations, angles of attack and yaw
attitudes. Several examples of the correlation achieved with
test data are presented in Appendix B, Figures B-3 to B-5.
Both the APT and potential flow program have been demonstrated
to be extremely useful tools in the design and analysis of
rotorcraft configurations.
28
METHOD APPLICATIOIT MID CORRELATION
The hub drag prediction methods described previously require
information about the flow field in the pylon region and about
the ellipsoidal fairing if applicable. Because the drag
prediction method is relatively sensitive to the local flow
velocities in the hub region and the potential pressure
recovery indicated by ACpz, it is recommended that the partic-
ular pylon/fuselage configuration of interest (and the
ellipsoidal fairing, if applicable), be geometrically modeled
and an aerodynamic analysis perfornid to obtain the required
I
informatio.i.
Unfortunately, even with the simplified modeling technique,
this may require more effort than is warranted for a quick
drag estimate, particularly if the fuselage/pylon configuration
has not been finalized. To alleviate this difficulty, several
pylon configurations hav3 been analyzed in conjunction with a
basic fuselage representation as well as several general
ellipsoidal fairing configurations. An example of the geo-
metric model of one pylon configuration on the fuselage under-
body used is shown in Figure 16.
The basic pylon shapes used in the analysis are shown in
Figure 17 and are designated A, B and C. For each pylon shape,
the maximum pylon height-to-length ratio and the axial posi-
tion of maximum height were varied in order to obtain the
matrix of thirty-six general pylon configurations identified
of pylon from the potential
in Table
flow 1. The
program surface
(Y179L) the matrix obtained
for pressures shapes are shown
in Figures 18 to 20.
The ellipsoidal fairing shapes which were analyzed are shown
in Figure 21. The thickness-to-diameter ratios of the fairings 4
vary from .1 to .4. Predicted surface pressures along the
top centerline and the lateral centerline for the fairings are
shown in Figures 22 and 23, respectively.
In application, the user i3 required to identify the various
frontal areas associated with the hub configuration of
interest (i.e., Ap or AF, AS and ABS) as well as the hubi
position, shaft height, and the various emapirical drag
coefficients of individual components as needed. The partic-
ular pylon geometric characteristics such as length, width,
maximum height, and axial position of maximum height must be
kncwn. The values of the local surface pressures must be
determined in one of two ways. If the particular pylon/fuse-
lage or fairing configuration has been modeled and analyzed,
i
then these values are obtainel from the potential flow analysis.
The alternative to this is to identify a pylon or fairing
configuration from the ratrix identified in Table 1 and
Figure 21, respectively, which most closely represents the I
29
I
t4
u~~u. m. u. -.
0
20 0 0
,,4
0
0
z w
9n 0 I
30
''I
particular shape of interest. The values of Cpz, Cpt and/or
CPc, Cps can then be obtained from Figures 18 to 20 and
Figures 22 and 23. Interpolation may be required to obtain the
required pressure coefficients if the maximuum height and
position of maximum height or fairing thickness to diameter
ratio of the particular shape do not correspond to any of the
configurations available. Values of CPF for the particular
fairings shown in Figure 21 can be obtained from Figure 24.
A generalized hub drag prediction computer program (Y179Z) has
been developed which accepts the required geometric
characteristics of the pylon, hub and the shaft and blade
shanks (if applicable), along with the required empirical drag
coefficients. The predicted aerodynamic data for the pylon
matrix and the ellipsoidal fairings have been incorporated in
the program in order that an estimated pylon surface pressuze
distribution for an arbitrary pylon shape, and Cpc and Cps for
an arbitrary ellipsoidal fairing can be computed if required
in the hub drag prediction. The hub type is specified and
the appropriate equations (i.e., unfaired, ellipsoidal fairings,
rigid fairing) are solved to identify the drag of the hub
itself, the interference drag, and subsequently, the total
incremental drag of the rotor hub. The use of the hub drag
prediction computer program is described in Appendix C.
The procedure required for predicting the drag of a particular
hub/pylon/fuselage configuration is sur.unarized in the
following steps:
1. Determine Ap or AF and Aw, AS and ABS
2. Determine factors K2 and K3
3. Determine CDS, CDBS and Cf
4. Determine CpZ, Cpj Nnd Cpc, Cps, CpF (if applicable)
using the aerodynamic analysis or values from the
pylon and fairing configuration matrix
31
I.
* a 4% error in total aircraft drag of cr)nventional helicopters.
However, for exceptionally clean hel.4copters, an incremental
hub drag error of this size could represent about a 7% error
in aircraft drag.
* Some of the hub configurations identified in Table 2 were
tested with only a pylon and fuselage upper surface repre-
sentation and not a full fuselage component. The predicted
incremental drag coefficient values for these hubs (with the
effect of the fuselage included) are shown in Table 2 also.
The large differences between the drag values with and without
the fuselage present deserve further discussion.
The local pressures and velocities in the hub region are
dependent upon the full configuration. The hub configurations
tested without a fuselage weze tested with a pylon and fuselage
upper surfAce simulation similar to the configuration shown
in Figure 25. The local velocities along the pylon are ex-
pected to be lower for this type of configuration than for a
complete pylon/fuselage shape. Although no test data was
identified during the course of this study which substantiates
this conclusion, the differences in the calculated surface
pressures for a pylon/fuselage comb~ination and the same pylon
with only a simulated fuselage upper surface are shown in
Figure 26. The difference between the pressures is essentially
equal to the surface pressures along the top of the fuselage
body without the pylon or, in essence, the fuselage surface
pressures are superimposed on the pylon surface pressures
except at the end of the pylon where the pressure differences
are small.
The predicted hub incremental drags for the hub configurations
tested with a pylon and fuselage upper surface are based on
the pressure distributions on a pylon/fuselage upper surface
configuration, and good correlation with test data is obtained.
Since good correlation was achieved with the configurati.ons
tested with a full fuzelage component and those without a full
fuselage, provided the appropriate local pressures are used,
it is reasonable to expect that the predicted incremental hub
drags based on the complete pylon/fuselage configurations
shown in Table 2 are accurate.
The differences in the predicted incremental drags with and
without a fuselage component demonstrate that the incremental
drag of ellipsoidal faired hubs is quite sensitive to the fuse-
lage presence and that the drag of rigid fairings, and probably
unfaired hubs, are not nearly so sensitive to the presence of
the fuselage. This tends to further support the earlier con-
clusion that the interference effects associated with ellip-
soidal fairings are much greater than for the other hub
configurations.
32
rr
44)
c. - H C' U E3 . 0 :
F4)o 0 4 S + +. I 4. I
E-4 H I- H0r
b)
Cu 4)3
E-
;4
04
@1 H0H91- H H H w.4
43 4.. 41
0. Hd i $4.2
0 0A 0. t- 0
V 0 4 U.- a .- ' 0 - V4 3
C) 4) . 4A~ 0 4 - (fl 4) (( 1 0 N D
V0-a W1 - . H *. o c r. o@
0- H
%D 0H) HD H 0 4 1 4) .
0.. 4 04 I 4,4
0)CO )
0) 4 D0 Q0 4) *
F*z 4. 40 0.-1 H4 "44 rq o
4)
ca@1)
0) 4) f1 4) H H H H 4)P
a. H H H H H bO
E-irz. z z ~~ ~-4)
33)
HUB FAIRING DESIGN
34
WIND TUNNEL TEST DEFINITION
* 35
Recommended Data:
a. Wake survey
b. Laser velocimeter
5. Surface pressures
36
/
77 ~~~ ~
1111 ~ 27" ' .1 .,7
Model Scale:
Reynolds number based an hub diameter should be greater
than 1.5 million. This yields a minimum diameter of
approximately .8 feet based on the wind tunnel des-
cribed above. If the scale is based on the configura-
tion shown in Figure 27, the minimum scale is one-sixth,
which yieldsl low tunnel blockage factor of approxi-
mately .17%.l
In light of the results presented in the previous section, the
wind tunnel test program should attempt to verify the validity
of rotor hub drag tests without a full fuselage sinulation.
Even if this approach is dervnstrated to yield lower valur~s
for the hub incremental drag than with a complete fuselage
underbody, this does not preclude the possibility of obtaining
meaningful data provided that the mechanism responsible for
the differences in hub drag magnitude is identified and
understood in greater detail.
Several pylon configurations are proposed for the test programI
to permit study of how the pylon shape itsel~f affects the hub
incremental drag. In addition, a matrix of pylon configura-
tions such as the one proposed will provide a better data base
for the design of low drag pylon configurations. Testing
* selected pylon configurations with the simulated fuselage upper
surface as well as with the comnplete fuselage underbody would
help better identify the interference effects associated with
a pylon/fuselage configuration.
The axial location of the unfaired rotor hub should be variedJ
-for each pylon shape to aid in refining and verifying the hub
drag prediction method. The positions tested should be
* ~located forward of the maximum height position or at theI
maximum height position. The basic hub position should allow
the hub to be rotated, but rotating the hub for the hub
position study is not considered necessary.
A component buildup (i.e., control rods, etc.) should be per-I
formed for the basic hub configuration and a systematic study
of the effects of shaft heivht should be done at the basic
axial hub position (i.e., the position at which the hub balance
is located). It is not considered particularly necessary to
rotate t~he rotor hub for the basic hub position and height
studies, but the hub should be rotated to verify this.
A representative set of rotor blades should be included in j
this study as well. No data were found in the literature
which demonstrate how the rotor blades and the rotor wake may
affect the way in which the presence of the hub interferes with
4A
Pope, A. and Harper, J. J., LOW-SPEED W111D TUNNEL
TESTING, John Wiley and Sons, Inc., New York, 1966.
37
the pylon/fuselage configuration. One would suspect that the
rotor wake dominates the flow field in the hub/pylon region
and that the mechanism of the interference drag is significant-
ly altered. The surface pressure data and the oil flow studies
will aid in idenlifyin( the rotor wake effects on the pylon/
fuselage region.
The rotor hub has additional undesirable effects on the
helicopter overal, performance other than increasing the
parasite drag. The wakc from the hub can interfere with the
tail surfaces and the tail rotor, causing undesirable handling
qualities and vibration. Reducing the -rag of the rotor hub
should reduce the magnitude of this interference, and properly
designed oylon shapes can alleviate the possibility of this
interference by altering the wake trajectory. For this reason,
a wake survey is reconmended for the faired and unfaired hubs
on the three types of pylons to investigate the potential
reduction in the hub wake interference effects.
Laser velocimeter measurements are also -.aconmended in order
to obtain oif-body velocities both with and without the rotor
wake present. This will not only aid in defining the region
of influence of Uiduced velocities due to the fuselage/pylon
configuration, but also aid in determining the unsteady flow
characteristics associated with the rotor wake effects and the
rotor hub wake.
Although some of the test program characteristics are not
inherently necessary to refine and verify the hub drag pre-
diction method, these features will significantly increase the
understanding of the entire hub/pylon/fuselage flow character-
istics. This will aid in the design of efficient high-speed
rotorcraft. This test program also offers the potential to
provide a data base directly applicable to two cont*acts
recently awarded by the U. S. Army Air Mobility Research and
Development Laboratory, Eustis Directorate, concerning three-
Jimensional aerodynamic analysis techniques in extensive use
to predict flow velocity and direction off the sur-face of the
body. This analytical capability is essential for predicting
flow interference on external stores and identifying potential
body/rotor interference.
The wind tunnel test outlined is rather extensive in scope, and
consequently, the estimated cost of the program is high
($415,000). This cost is comprised of approximately $112,000
for the design and fabrication of the model, $44,000 for the
design and fabrication of the rotor hub (assuming a properly
scaled hub is not available), and $259,000 for the wind
tunnel program and documentation. The cost estimate is based
on the use of existing balance systezr. and rotor blades
compatible with the test outlined.
38
"CONC LUSIONS
Results of the helicopter rotor hub drag data review indicate
that:
1. The rotor hub swept frontal area is the primary
factor influencing the contribution of the rotor
hub to the total aircraft, drag. Reducing the
frontal area orfers the greatest potential to
reduce the hub drag.
2. Changing fuselage angle of attack does not
sign~ificantly affect the incremental drag of
unfairc~d or faired hubs,
3. The effect of rotation on the incremental drag of
unfat.red hubs, having four or more blades, is
negligible. The data reviewed on the effects of
rotation on faired hubs was inconsistent and
no conclusions were drawn.
4. Increasing the separation distance between the
rotor hub and the pylon can reduce the incremental
drag of the rotor hub only if the drag of the
increased exposed shaft and control rods is less
than the reduction in interference drag caused
by removing the hub from the higher dynamic
pressure region around the pylon and the resultant
smaller flow disturbance on the aft pylon/fuselage
region. This conclusion is based on test data
obtained without the effects of the rotor wake
present. Due to insufficient data, no conclusions
have been drawn concerning the effects of increased
shaft height on the airframe/rotor mutual inter-
ference.
5. Small-scale tests can yield accurate hub drag
values only if the actual rotor hub geometric
properties are properly modeled.
6. Rotor hub fairings offer the potential to yield
signi.ficant hub incremental drag reductions. For
ellipsoidal type fairings, care should be taken
to guarantee that the fairing causes a minimal
increase in frontal area. This is particularly
important if boundary layer control is not used
in conjunction with the fairing.
7.The drag reductions available using rotor hub
fairings are decreased as flight Mach number is
increa ed due probably to local shocks occurring
on the bluff aerodynamic shape of the fairings.
39
8. The rotor hub and its associated interference
effects contribute 20 to 30% of the total airframe
drag of current helicopters. If the magnitude
of increnental hub drag is not reduced, the laub
could potentially contribute 40 to 60% of the
airframe drag for aerodynamically clean fuselage
configurations.
40
REFERENCES
41
11. Hess, J. L. anw Smith, A.M.O., CALCULATION OF NON-LIFTING
POTENTIAL FLOW ABOUT ARBITRARY THREE-DIMEW4SIONAL BODIES,
Journal of Ship Research, Vol. 8, No. 2, 1964.
42
BIBLIOGRAPHY
11. Olson, J. R., WIND TUNNEL DATA AND ANALYSIS OF FULL SCALE
ROTOR HEAD AND PYLON DRAG INVESTIGATION, SER-61027,
Sikorsky Aircraft, April, 1959.
44
0
0 N0
oo 0
N C)
~0
0 0a:
'.0 - S)E
cI u L.
o wU
Uo
02 U0
In 0
~ ~
'OV80~~ ~ -lI3430Ien
00
U,45
50r-
L FUSELAGE DRAG
REDUCED BY = 50%
S(HUB DRAG HELD CONSTANT)
40
I-- 30 CS-65
w CIýL$ 0 CH- 47C (I HUE)
w ,: S-550 S-58 0 S-61L
0 S-6200 S-667
Z 20, bO-1050 S-560
no ,0S-64
.r.=
0 10
o 0 REF IT
0i
0.I if I 1 liml
2 4 6 10 20 40 60 100 200
AIRCRAFT GROSS WEIGHT, 1000 LBS
46
//
20 FLASGED SYMBOLS INCLUDE
BLADE SN1ANKS NOT
ACCOUNTED FOR IN Ap
16
N
c"
Ir-
: 12
4 - C REF. 18
0 REF. 19
M REF. 3
0 4 8 12 16 20
2
HUB FRONTAL AREA, FT
Figure 3. Hub Drag Trend With Hub Frontal Area for Unfaired Hubs.
47
20 FLAGGED SYMBOLS INCLUDE
CU BLADE SHANKS NOT /
I.- ACCOUNTED FOR IN Ap /
, 16 -'
or /
r C1 SUMIN LOCALq
-, /0 fM ASSUMING - 1.25
0f
W
z
w
8
op(.582 +.0349Ap2
zi ~
4 ,q/ 0 0 REF. 19
REF.4
D' REF.3
R
0 REF.7
0 4 8 12 16 20
HUB FRONTAL AREA, FT 2
Figure 4. Hub Incremental Drag Trend With Hub Frontal Area for
Unfaired Hubs.
48
z 4z
oL
CD
zz
L<LL
00
Li.L
w ai
z 49
C,1
LLLLLL - (l a
00 00
0 CA
-0. L&.;
Z 0r
'a4-
-J U- 4
w IY at~ a)
CLtil
Y38)
'I-N lD0UV n LL.-
soLD
L 1 J-J
_ 3
c- C) L .t
tun)2 04-)
UEU
C) LJU
, 0 0.
0i Ct
0 L0
0.
o~C u. -
0
Id
LIa.
InI
0 U. 4-)
N
~IOL3V. OVO NOAd-SlI-
!>i J33~3~31N
51 so
ui LLI
0D -E
D LL
M o1
4U 0
LJ
UUI
0 1;
LLL
CO N 0
~1V .H N3VJd~N4-)l
'D~G
z (U
52.
24-
L 0 REF 3 (M .3)
6 REF 18
20
•• •• "• j•ELLIPSOIDAL FAIRING
20-
SW/O BLC
ARTICULATED
N
&u. 16--
16
m" RIGID
'12.
LiL
z .FULL SCALE
.• ARTICULATED
FULL SCALE
4 [ } •, E3FULL SCALE E
ARTICULATED
-A AFULL SCALE
DIRECT TILTING
0 20 40 60 80 I 00
ROTATIONAL SPEED, 96 MAX RPM
53
40- DATA FROM REF 3
324
RIGID FAIRING
z
S16-
U
Z W/BLC
*/
D L8 ELLIPSOID FAIRING
INCLUDES PENALTY)
(
FOR BLC POWER
"'I I I I "t-- I
0 .1 .2 .3 .4 .5 .6
MACH NUMBER
54
OPEN SYMBOLS - FULL- SCALE TESTS (Re - 6.5 X 106)
SOLID SYMBOLS -I/6 SCALE TESTS (Ret 1.OX i0)
20-
16- UNFAIRED
! 6, HUB
fVP
0 IFAIRED
- HUB
PF
4
4 . .. P
0 I I I
-12 -8 -4 0 4 8
ANGLE OF ATTACK, DEG
Figure 11. Comparison of Full-Scale and One-Sixth Scale Rotor Hub Drag
Test Results.
55
W NX
'%4J
CLC
Z.8c
C3
0.2
'U
OD 0 Ow
0 0 LL
'IN104400ovaa NOlIkd
56
C
0a
4-I
c
0
41
CL
#a41
Pd ~0
Lai
L .0
57
0
ON
.44
5.4
.9:4
-V
5- 0I
x r4
'WCW
0-
4.c0
ON
-r4
0
,4 .
LL
t -44L
• •:C
0o
I.
!a
/0
r 4-
/U
/i/..- "N
•.
CJ
600
Lir
°C
* ',,--*----.--. -- - -- ~.
zz
a: I
to
4J I
4 4 II
I 2 ~ ~~PYL.ON
TYPE A-MAXH *7 ;5o% -axNa'O
•
!.,PYLON TYPEt M T
+
S5%
0 10%'A
4 20%-
-T6
0 20 40 60 s0 100 0 20 40 60 sO t00
PERCENT PYLON LENGTH PERCENT PYLON LENGTH
t --
, * I '
4
S
0 20 40 60 SC O00
PE[RCENT PY'LON LENGTH
i2i
62
S/ ____
mix I
PYLON TYPE IMAX.t ATI
MAI
IPY
I I
TYPE , -MAX.
I I
%
0 20 40 60 so 00 O0 20 40 60 sO t00
PERMNT PVLON LENGTH PERCENT PYLON LENGTH
• .41
-. 4 A-
0 20 40 so so 100
PERCENT PYLON LENGTH
63
/I
/
1.2
YLNYPc
C . MAX. H AT 25% PLNTP AHA 0
.8 MAX.N
p i i IA•
.. 4
0 20 40 60 so O000 20 40 60 so too
PERCENT PYLON LENGTH
PERCENT PYLON LENGTH
.8-
PYLON TYPE c - MAX. . AT 75 %
-.4 . .
.4
..a
64
t/d
-.4
-.3-
100%
DIAMETER
65
.i
i/
_e 0
I.*J4 0 P4
4r41
I-d
D44
66
t m
- p
-c -- -r4 - fn :3
440
Iu 0
L-- 1. 0~
dI IinS38 -:vi
670
00
0 ,
0wIL• -. 02
"'-.06
0 .I .2 .3 .4 .5
HUB FAIRING . t/d
68
Figure 25. Geometric Model of a Pylon/Simulated Fuselage Upper Surface
Configuration.
0* b:
#4
* I / ' L AGE
L0 20 40 Go 00 100
PERCE•NT PYLON¢ LINGT"
69
"?.
fro"
00
44
M C3
Im u
.. 1
70
APPENDIX A
GENEPAL DESCRIPTION
The Automated Paneling Technique (APT; Program Designation
Y179A) was developed to simplify the generation of a geometric
model of any arbitrary shape compatible with Sikorsky Aircraft's
three-dimensional potential flow aerodynamic program (Y179L).
The program accepts inputs from the user which generally des-
cribe cross sections of the body along an axis by combinations
of curves and straight line segments. The cross sections are
ý.ncremented by the program based on specific constraints so
that each segment, and consequently each cross section, is
approximated by a series of straight line increments. The
various increments on adjacent cross sections along the
axis are then connected to form quadrilateral panels which input
describe the surface of the body as required by the aero-
dynamic technique. The generation of a proper geometric body
is simplified by reducing the user's responsibility for incre-
menting each cross section and properly identifying every panel
and its corner or nodal points. Because of the constraints
placed on the incrementing process, the panels generated by
the program should approximate the minimum number of panels
required to properly iiodel the surface of the body. Figure
A-1 illustrates the type of body model generated and the
coordinate- system used.
71
/
3. An increment length may not be more than 130% or
less than 70% of the length of increments adjacent
to it,
4. An increment must approximate the true curve by
a specified tolerance.
5. The total number of straight line increments must
equal the nursber of increments at the cross sbction
to which this section is connected to form the
quadrilateral panels.
6. User specified starting or ending increment lengths.
A priority is associated with each constraint and consequently
all sections generated nay not satisfy every constraint. In
this respect, certain constraints represent only approximate
restrictions. For instance, for some configurations, the
relative size of adjacent increments may not strictly adhere
to constraint 3 due to an inability to satisfy all the con-
straints which have been specified.
Any number of segments may be used to describe a given cross
section, but generally the more segments used, the more
likely it is that the program will not generate a proper model
of the body. For most cross sections, two or three segments
are sufficient. The points that must be specified by the user
for each segment type are shown in Figure A-2. The open symbols
define tangency conditions only and the solid symbols repre-
sent actual surface points. A single point is an acceptable
input, but cannot be used in conjunction with any other
segment. If a single point is used, the program assumes that
the section described consists of only that point, and it is
normally used to descrJbe the nose of the body. An example
of an input section and the resulting output section generated
by APT is shown in Figure A-3.
The APT progran assumes the body to be 3ymietric about the
Y-Z plane and consequently, only the half of the body in the
positive X direction is acceptable as input. The program will
automatically perforri linear interpolation to generate inter-
mediate body sections along the input axis as requested. This
feature allows the user to control panel length without re-
quiring additional input sections.
In addition to the general features already described, APT
incorporates several user oriented options. Some options
are designed to give the user more control over the distribu-
tion of surface panels generated, others are used to input
sections which do not conform well with the standard input
procedures. Plot options are also available to aid in gener-
ating a proper geometric model of the body.
72
The 3tandard input procec ure requires that cross sections
along the body axis (" axis) be described. Often a body or
sections of a body would be r.ore easily described and conform
to the incrementing process better if cross sections were
described along the "X" or *Y" axis. Fuselage sponsons,
rotor head fairings and wings are good examples of these types
of sections. The program will allow the user to describe
these'sections along the X or Y axis as requested. If this
option is selected, the cross sections must be input from the
maximum "X" or "Y" value to the minimum instead of the stan-
dard miniimum to maximum "Z" values. The proper order for
inputting segments at any station along the input axis is shown
in Figure A-4.
The program marches along the requested input axis increnent.tng
each section. The section at any one station may be described
by a different series of straight line increm-ents, depending on
the sections to which it is connected. That is, the number of
increments required to connect the section to the previous
section may differ from the number of increments required for
the section following it. Two sections can be described at
one location on the input axis. In this case, the program
assumes that the body has a discontinuity at this station
and a message stating that normal panels are required will be
written (this message will also appear in the punched card
output generated). Defining two cross sections at any one
location of the input axis is required only when the cross
sections to be modeled are of different shapes (for instance,
the location of an engine intake or exhaust). Two cross
section definitions are not required to alter the number of
panels describing the surface upstream or downstream of an
input section. The program will not generate panels normal
to the input axis and it is the user's responsibility to iden-
tify normal panels if a closed body is to be generated.
For basic fuselage shapes without a pylon, wings, or engine
nacelles, the body is modeled in one piece. However, portions
of a body may be more complex and require several more panels
in regions to obtain a proper geometric model. In this case,
the more complex regions of a body may be modeled independently
and the resulting output combined to form the complete con-
figuration. This is demonstrated in Figure A-5. The Advancing
Blade Concept aircraft (ABCTM) main rotor pylon and the engine
nacelles were modeled separately for two different reasons.
The main rotor pylon required several more input stations along
the axis than were required for the fuselage in the sane region.
Consequently, in order to reduce the number of panels required
to model the configuration, the two were modeled separately. A
model of the ABC with the pylon was generated prior to the
decision to add the engine nacelle models. In order to add the
nacelle, the panels describing the fuselage in that region were
73
simply removed from the data deck and the panels required to
interface the nacelle (which was modeled by APT) were sub-
stituted. The panels at the wing, pylon, or nacelle inter-
section with the body must be inputted by the user. The
numbering sequence required for the nodal points is shown in
Figure A-6. A separate program (Y179C) is used to obtain plots
of the final combined configuration.
HELPFUL HlNTS
Although the automated paneling technique reduces and sinpli-
fies the user's responsibility, care must be taken to insure
that the body shape iL modeled correctly. The printed output
from the program is designed to aid the user in identifying
improperly modeled cross sections and nonplanar panels (i.e.,
surface panels which do not adequately approximate a flat
surface). The plotted output also aids in this task. This
section is designed to assist the user in proper use of Y179A.
Because of the numerous conrtraints placed on the incrementing
process, certain combinations of segments used in describing
a cross section may crct.e an improperly modeled section. The
user should also insure that individual segments are described
properly.
Experience has shown that two errors in Liputting circular and
super-ellipse segrfents occur frequently. The error which
occurs most often in describing a circular element is a result
of unequal radial distances from the center of the circle to
the first and last coordinate points of the circular segment.
If the two radial distances vary by more than 5%, the segment
inputs are identified as being incompatible and a message to
that effect is printed in the output.
The most common error occurring in describing a super-ellipse
segment occurs when the third input point does not lie in the
triangle shown in Figure A-2. Most often this occurs because
of improperly defined tangency lines. If the third point lies
outside the given triangle, the segment inputs are identified
as being incompatible and this message appears in the output.
Both of the errors listed above will cause the program to omit
the cross section at which the error occurs and continue. How-
ever, the program will terminate prior to generating the sur-
face panels.
Another error which sometimes occurs in describing the cross
section is that of inputting a small straight line segment
adjacent to a circular or super-ellipse segment having small
curvature. This condition may cause the increments at the seg-
ment intersection point to fail the ±30% size relationship.
This will not cause the program to terminate but may cause an
unsatisfact-ory increnent d.strDiution.
74
I
I
The most common problem encountered in the surface modeling is
nonplanar panels generated by the program. If the sections
used to form the surface panels are similar in cross section
and have no discontinuities in the curvature, this problem is
seldom encountered. It is the user's responsibility to insure
that the sections used are similar. Often sections used in
forming surface panels are quite similar but, as mentioned,
discontinuities in curvature are present. This condition will
sometimes cause nonplanar panels to be generated.
The program incorporates a planar panel check which will iden-
tify what panels fail to satisfy the planar criteria and by
what percentage they are nonplanar. The user is responsible
for modifying the nonplanar panels if required. This can
usually be accomplished by simply changing the quadrilateral
panel to two triangular panels. Changing the specified surface
maximum and/or minimum often eliminates nonplanar panels and
it is recormended that these constraints be varied if non-
planar panels are a problem. Experience has shown that panels
within 9 or 10% out of plane are sufficiently accurate for use
in Y179L. In addition, nonplanar panels nay be identified in
regions of the body not of particular interest to the user or
in regions where the assumption of potential flow is not valid.
In this instance the user may consider the panel representation
su,.ficien tly accurate.
Despite the built-in safeguards, the automated paneling tech-
nique iS not fail-safe ad-a gradual buildup in the complexity
of the bodies-m6215110 is recommended.
IN1PUT .IUSTRUCTIOUS FOR Y179A
The Fuselage Geometry Definition Program requires punched card
input. The input procedure for Y179A is designed to simplify
the generation of a data deck describing a general fuselage or
body shape for input to the Three-Dimensional Potential Flow
Program (Y179L).
DESCRIPTION OF CARD INPUT
Card No. Columns Symbol Description
1 1-72 TITLE Title card - Any alphanumeric
characters acceptable. This
title will be used to identify
all output generated.
2 1- 4 TYPLT Plotting unit to be used if
plots are requested. (left
justified)
BLP - Calcomp Plots
Blank - No plots generated
8 Enter "l" if sideview plot is
desired
75
bI
Card No. Columns Symbol Description
9 Enter
are "1" if cross section plots
desired
76
Card No. Columns Symbol Description
15-21 DSMI Enter the minimum increment
length acceptable for this body.
This value is subject to change
by later input.
22-28 DSIfl Enter maximum increment length
acceptable for this body. This
value is subject to change by
later input and/or program re-
quirements.
29-35 EF By entering any nun r >0.0 the
user may expand or cuntract the
input coordinates by the value
of the number entered.
36-42 XOVR Any number >0.0 in this loca-
tion will override the maximum
increment length criteria based
on the separation distance of
opposing increments.
78
ii.
8-14 X1 Xl
15-21 YL Y1
22-28 X2 X2
29-35 Y2 Y2
36-42 X3 X3 - If a specific nurber of in-
crements is desired for a
straight line segment, ernter t.'e
number desired.
"43-49 Y3 Y3
50-56 X4 X4
57-63 Y4 Y4
64-70 X5 X5
71-77 Y5 Y5
78-80 MIEQ A negative integer entered will
cause the length of the first
increment of that segment to
approximate the product of the
integer and t he minimum surface
distance specified.
79
/ 1
The required cards for the next body station along the input
axis follow the last segment card (starting with card IS 1).
If the body station entered is equal to the previous body
station, the program assumes that this station has been described
by two different cross sections and that additional input panels
will be required at that station.{
A blank card should follow the last body station segment cards
if another body shape is to be generated. If no other shape is
required enter "END" in column 1-3.
Required CPU time for a typical 400 panel configuration is ap-
proximately 1 minute.
80
INPUT INSTRUCTIONS FOR Y179C
Y179C is an auxiliary three-dimenuional body plotting routine.
Its purpose is to enable the user to obtain plots of a body
shape u3ing card output generated by Y179A.
The card output from Y179A which describes the surface panels
for an arbitrary body m.ay be modified and additions to the
body may be incorporated. Isometric, cross section, and/or
side view plots of the resulting body may then be obtained by
using Y179C.
The following cards are required for every body shape to be
plotted:
Card Columns Description
1 1-78 Title Card - Any alphanumeric
data is acceptable.
2 6-10 Total number of panels des-
cribing the body shape (N -
integer, right justified).
Format for Panel nodal points:
Card No. Panel No.
3 1 Node 1 (XY,Z) Node 2 (X,Y,Z)
4 1 Node 3 (X,Y,Z) Node 4 (X,Y,Z)
81
The following card is used to identify a coordinate system off
set for various panels if required. For instance, if the input
panels represent mlultiple bodies, then the separation distance
between bodies may be changed.
Card No. Columns Description
2:1+6 1- 5 Panel number "NID" (Integer-
right justified).
6-10 Panel number "NJE" (Integer-
right justified).
11-20 X17 These values will be
21-30 Yq added to all panels
31-40 ZV between panels IM and 14E.
82
0
I-f
Co S
00
z 0
•-- w. 1
Sp-
83S
Z w)
cn w
00
ZW
00
010
CY'
Snn
w - 0
I--
- *84
-T--
WI 0
* 44
V)
04
zz I
0rw
o~A0
U))
z¶
85Q 0
FRONT VIEW
(Z TO Z OPTION)
+ Y
+X -x
TOP VIEW
Y TO Z OPTION)
+Z
+X -x
SIDE VIEW
+ y X TO Z OPTION)
+ z-z
06
4.'
C;
0
I44
4-I
U)
0
44'
87-
)3
i4 (TRIANGULAR PANELS f
2
4 3
, N (QUADRILATERAL PANELS)
7 N
92x ik
IS IN THE SURFACE OUTWARD NORMAL
VECTOR DIRECTION.
88
APPENDIX B
DESCRIPTION AllD USE OF THE TIIREE-DIIMSIONAL AERODYNAMIC
ANALYSIS TECHNIQUE-PROGRAM Y179L
GENERAL DESCRIPTION
The analytical technique used to predict the aerodynamic flow
field on and about an arbitrary lifting or nonlifting con-
figuration is described in more detail in Reference 9. The
basic potential flow methods used in the technique are des-
cribed in detail in References 10 through 12.
The technique requires that the surface of the body be des-
cribed by a finite nunber of approximately flat panels as
shown in Figure B-1. Any lifting sections such as wings and
tail surfaces are also modeled by a vortex lattice system on
the mean carter surface as shown in rigure D-2. A constant
source strength is distributed uniforn.lly over the surface of
each surface panel and the vortex lattice orystem consists of
bound and trailing line vortices. ThL geraral solution of the
flow field is then obtained by -D!ving a set of linear
simultaneous equations relating tha •:fluenca of each source
distribution and line vortex on all the other surface panels,
control points and the points at wihich the Kutta condition is
applied for each lifting section. The solution of these
equations yields the source and vortex strengths from which
the flow velocities and directions about the body or bodies
can be calculated.
"The aerodynamic program accepts inputs from the user de-
scribing the surface panels by their corner or nodal points.
The order of input is arbitrary since each panel is described
independently. This feature also allcis the aerodynmaic
solution of multiple bodies. The technique accepts triangular
as well as quadrilateral panels.
The vortex lattice system for lifting sections is generated by
the program, given the spanwise position of trailinq vortices
and the requested chordwise position of the bound vorticity.
The particular vortex lattice technique used requires that the
chordwise distribution of vorticity be known. A modified
.Iulthopp lifting surface analysis is employed to determine the
vorticity distribution. This feature is inherent in the
program and elininates the requirement for specifying the
vorticity distribution. This method is most accurate when
applied to thin, high aspect ratio wings and is less accurate
for lifting sections that are very thick or have a snall
aspect ratio. It is, however, considered sufficient for even
these shapes since only the chordwise vorticity distribution
89
is determined by this method and not the vorticity strengths.
The program is capable of handling arbitrary configurations at
angles of attack and/or in yawed flow. Engine inflow or ex-
haust can be simulated by requesting fluid ejection or suction
through the surface of individual panels. In addition to these
features, the program incorporates several user oriented op-
tions for rotating the bodies or coo.'..:-'iate offsets. These
features allow the user to vary winqit%4dy incidence, wincj
dihedral, and fuselage/stores separation distance without
generating additional body descriptions.
90
.. a---t..lii" ... .....-. f+-
91
U6
92
---- - -
Card Columns Symbol Description I
5 5 NSAV1 0 - Do not file geometry data
-1 - Geometry data not on file -
initiate file write of
geometry data (TChis feature
is used to store various
geometric dcaracteristics '
to avoid time-consiuiing com-
putations if subsequent runs
are anticipated for this
sha,) .
1 - Get'.,' try data on file -
geometry data is to be read
from file
10 IJSAV2 0 - Do not file solution
-1 - Solutions to various flow
conditions are not on file
- initiate file write of
solution
1 -Solutions for various flow
conditions are on file -
search file for required
soluticn - calculate
solution and write on file
if solution is not currently
on file.
Various sections of the body or bodies may be inputted indepen-
dently if desired. The fo'lowing cards are required for each
individual section used (maxinum of five sections).
6 1- 5 TIYPfl Body Section type:
:IL,S - nonlifting, symetric
:JL,NS - nonliftiny, nonsym-metric
L,S - lifting, syl.mietric
L,ZIS - lifting, nonsyr.z.i•etric
13-72 Any alphanumeric title de- 4
scribing this section
7 5-10 "iop Total number of surface panels
describing this subsection.
(Ii the section type on card
6 designates a nonsyirnetric
section, the body is assez.ied to
be in t'o subsections, one on
either side of the Y-Z plane).
8 1-10 XOF Location of origin in section
11-20 YOr coordinate system. Trhese values
21-30 ZOF will be added to the panel
coordinates to obtain coordin-
ates in reference system
93
Card Columns Syrbol Description
94
V
Card Colunins Syr•bo 1cscription
1- 5
12 NIP Panel ntuiber for which rota-
tions are required. A zero
will exit to card 13. (Integer-
right justified.)
6-10
11-15 U•OD 1
NOD 2 Panel node points which should
16-20 NOD 3 be rotated. (Integer-right
21-25 110d4 justified.)
If specific points on the body are to be rotated,
repeat the following card for each desired point.
R3 1-10 Al (X, Y, Z) coordinates of
11-20 01 point to be rotated.
21-30 C1
95
Card Columns Symbol Description
Li 5 IV Body section number requiring
lifting section Inputs.
10 LV Subsection 1 or 2-
96
The followinrg cards designate various output options and will
be preceded by an "0".
01 1- 5 M Number of waterlines (Y -
constant) along which surface
pressures are to be calculated s
by interpolation between panel
control points. (Maximum 16,
Integer-right Justified.)
97
IV M ,0 repeat the following card as required:
98
Unit(s) Description
99
v
*14
pa
>14
'0
1.4 Eq
r4)
0-4
1001
41~
\144
LL.r
04
$4
-45
b4
C144
101
f, ap a. 1 psi,0 o-4 9
. . . . . . .t
. ......
- ..
A C L~ . . ... .. ... .~ . . . . ,
0 40 00 120 ti 0 Im
P16 40 2110 32 3G0 4100 ,10 400 WOSW05
WOOVSTAT10%I, *eINCHE
I.I
Model.
. . 01.20..
Surface"rCALCULATEDUsPoERRSJRAOI
- r EXPCRINTAL
UPP3ER SUAF
I
. ... 0.
#A.
IEXPERMNA~7
ITCE
_EIIM TL OWEN
S - ~CA4CUATED LOWER SURFACIE
0 20 40 60 so 100
PERCENT CHORD
102
-I.t4
00IOfA6A 9
* ~, . , *~* [y +-f j
~
0 _______7-
AS,.i~
o 10 o 3 0
6O SO 60 7o 0 90 1
bOOY STATION 'L LING?,,
103
24 .. I-I . .... .. ..
C, , I,
INITIAL RUN
ou (NO. ODl
PNL CPU IO-- i I
1-7----LonIBM30---S--y----
-TTAE.
B-6 Tim
e'o
INUMBELRU --
OF PANELS:
0. L USQET RUN
. ,/ L "'• : iUSING STOHREDu DATA
' - S--
;- ,"- .----- . .•. . -. - - • - -
4- 4
- _
- -
010
Figure B-6. Estimated Cpu Time for Y179L on 13t4 360 System.
104
APPEIDII% C
PROGRAM - Y179Z
(.C::LAL DESCRIPTION
105
Card Columns Symbol Description
4 1-.0 SC(i) Shaft height
11-20 SC(2) Shaft diameter
21-30 SC(3) Shaft frontal area (As)(sq ft)
31-40 SC(4) Shaft drag coefficient (CDs)
Non-zero values on Card 4 are required for unfaired hubs only
if the shaft height is considered sufficient to remove the hub
from the local increased velocity region due to the pylon.
5 1-10 BSC(l) Blade shank length
11-20 BSC(2) Blade shank frontal area/
shank. (ABS) (sq ft)
21-30 BSC(3) Blade shank drag coefficient
(CDBS)
, 10
106
LIST OF SYMJOLS
A projected frontal area, sq ft
CD drag coefficient referenced to frontal area
Cf skin friction coefficinnt
Cp pressure cce fficient,
d diameter
f equivalent flat plate area, drag/dynamic pressure,
sq ft
h pylon maximum height, in
t pylon length, in
14 Mach nunber
q dynamic pressure, sq ft
Re Reynolds number based on hub diameter (unless
otherwise stated)
t hub thickness, in
Z distance along pylon length, in
Subscripts
107 204-76
/'